What is the standard for evidence in biology?

Specifically, what is the evidence for common descent?(Not quite) famously, Darwin mused about the similarities of taxonomic hierarchies in linguistics and biology and asserted that the hierarchies must ultimately point to common descent. (Chapter XIV, On the Origin of Species) That’s common descent as distinguished from microevolution.

The linguistic equivalent is the single origin of all languages (eminently unproven and deemed unprovable) as distinguished from a language family (with demonstrable relevant organic shared features).

Darwinists are welcome to present their evidence. From Rumraket, we have the observation that all organisms can reproduce, “Nesting hierarchies are evidence of common descent if you know that the entities sorted into hierarchies can reproduce themselves. And that particular fact is true of all living organisms.” Good start.

From Joe Felsenstein we have the doubt that the border between micro- and macroevolution can be determined, “OK, so for you the boundary between Macro/Micro is somewhere above the species level. How far above? Could all sparrows be the same “kind”? All birds?” Not very promising.

From Alan Fox, “Darwin predicted heritable traits. Later discoveries confirmed his prediction.” Questions: Which heritable traits specifically? Was there a principled improvement over Mendel? And how does this lend credence to common descent?

Thanks to all contributors.

706 thoughts on “What is the standard for evidence in biology?

  1. colewd,

    This is the point Erik is making. You are trying to support common descent with similarities.

    If – a massive if – you couldn’t support common descent with similarities, you could hardly destroy it with differences now could you?

    Anyway it’s not just ‘similarities’, it is digital identity. Gene order, base order within genes, indels … 3 billion base pairs’ worth. I wish you information-jockeys would get your heads round the digital part of this argument.

  2. OMagain,

    What would you say to someone who said to you that they had read one book on linguistics, said they understood it (the field) perfectly and then said that the core accomplishments and understandings in the field were incorrect and unsupported, but without proposing something more fit to to replace them?

    When you try to force a hypothesis and there are no good alternatives you end up with bad hypothesis.

  3. Adapa: Like reading a book on the Wright Brothers then thinking you know enough to intelligently discuss modern jet aircraft design.

    Here’s a suggestion. Erik has said he is a linguist by profession. What about another thread, where we ignoramuses (linguistically speaking) read a book on linguistics (any recommendations, Erik?) and take on the professional. Would that be an analogous situation, Erik?

  4. Spaniards can’t communicate with Romanians, therefore both languages can’t share a common ancestor. It’s an observed barrier to language evolution

    And I’m an authority in linguistics because I once read a book that was written in a language

  5. dazz,

    Spaniards can’t communicate with Romanians […]

    Sound the same though. That’s good enough for me.

  6. dazz,

    Because You’re just a gullible darwinist

    Gotitinone!

    They do have a lexical similarity coefficient of 0.71 though. That’s science that is. What do we got? “These things look a bit alike”.

  7. colewd: When you try to force a hypothesis and there are no good alternatives you end up with bad hypothesis.

    Sure. But a bad hypothesis can be improved but there is nothing you can do with nothing.

  8. GlenDavidson: Yet you stick with ID…

    ID is a perfectly reasonable alternative to “it just happened, that’s all.”

    Your position is inherently a-rational. Things happen for no reason at all. So what is there to be skeptical of, exactly?

  9. @ Erik

    The evolution of plants tends to get overlooked. I have very much enjoyed Arthur Hunt’s contributions to the internet debate over the years. I was reminded of this by someone linking to this video which I enjoyed watching. I commend it to Erik.

    ETA given up on trying to make link work, see Dazz’s comment below. Thanks, Dazz. Encouraging to know at least one person follows my links!

  10. Alan Fox,

    The evolution of plants tends to get overlooked.

    Not to mention the fungi, algae, protists, prokaryotes … It’s all down to this curious obsession with chimps!

  11. OMagain,

    Sure. But a bad hypothesis can be improved but there is nothing you can do with nothing.

    A bad hypothesis can mislead people and create a lot of unnecessary work. The wrong hypothesis improved is the wrong hypothesis plus countless wasted resources. Erik’s point, I think, is that weak standards can lead you to collect a faulty hypothesis and now you’re basing your assumptions on the wrong conclusion, resulting in mounting garbage.

    At some point saying that we don’t have a good handle on what the data is telling us is a far superior conclusion.

    I would propose that the standard is the scientific method which includes direct testing of a hypothesis. If not, lets just admit we are having a philosophical discussion.

  12. colewd: A bad hypothesis can mislead people and create a lot of unnecessary work. The wrong hypothesis improved is the wrong hypothesis plus countless wasted resources. Erik’s point, I think, is that weak standards can lead you to collect a faulty hypothesis and now you’re basing your assumptions on the wrong conclusion, resulting in mounting garbage.

    Reality will intervene at a suitable point and point this out.

    colewd: At some point saying that we don’t have a good handle on what the data is telling us is a far superior conclusion.

    Your problem is the word “we”. There is no “we”. There is only you.

    The “we” that does not include you has already decided what the data points to as a conclusion. It may be wrong, it may be right. But if it’s wrong reality will intervene and demonstrate that at some point. It’s what always happens. Your current method of predicting the planet’s location does not work? Someone will notice that and find a better way. And so it goes.

    colewd: I would propose that the standard is the scientific method which includes direct testing of a hypothesis. If not, lets just admit we are having a philosophical discussion.

    How do we test the claims regarding Intelligent Design then? As no such claims have been tested, do you agree that ID talk is a philosophical discussion rather then science?

  13. colewd: A bad hypothesis can mislead people and create a lot of unnecessary work.

    And yet they will learn. They will learn why the current hypothesis is wrong, how the data differs from expected data if it were true and so on. All valuable information. You seem to be advocating sitting on your hands and doing nothing until the truth is magically revealed.

  14. colewd:
    OMagain,

    A bad hypothesis can mislead people and create a lot of unnecessary work.The wrong hypothesis improved is the wrong hypothesis plus countless wasted resources.Erik’s point, I think, is that weak standards can lead you to collect a faulty hypothesis and now you’re basing your assumptions on the wrong conclusion, resulting in mounting garbage.

    At some point saying that we don’t have a good handle on what the data is telling us is a far superior conclusion.

    I would propose that the standard is the scientific method which includes direct testing of a hypothesis.If not, lets just admit we are having a philosophical discussion.

    Behe, your hero, disagrees with you on CD. And at least he’s a biologist. What you propose is to throw away pretty much all the scientific knowledge and go back to the dark theocratic ages.

    You’re not making a scientific or philosophical case, just a lame political one.

  15. dazz: What you propose is to throw away pretty much all the scientific knowledge and go back to the dark theocratic ages.

    You really should get out more.

  16. colewd,

    At some point saying that we don’t have a good handle on what the data is telling us is a far superior conclusion.

    So apply that thinking to John’s ratite paper, or any other analysis of your choosing. You are saying, quite clearly, that there is NO good evidence, anywhere, of a genetic relationship between ANY species that cannot presently interbreed.

  17. John Harshman: Oh, come on. You know that they’re intended to demonstrate actual relationships among languages, and not just among languages with written ancestors. And you believe that many, though perhaps not all, of those trees are true. Do you actually think that the Algonquian languages have a common ancestor? Why or why not?

    The intention is not good enough. In historical linguistics, everybody draws trees. Everybody. But not all trees have the same value. Some trees are better than others. The important difference between those who stay within uncontroversial language families versus Nostraticists or pro-Altaicists is the standard they use – regular specific sound laws with irregularities explained versus broader typological similarities with looser cognate-detection and no attempt to explain the exceptions. The latter are more experimental and speculative. And more often demonstrably lax and wrong.

    So not all trees have the same value. This is painfully clear in linguistics. This is why the “we can draw a tree” argument is insufficient by itself. The tree must be built on solid standard, as a minimum, and better still, it should have independent evidence that it really represents what it is claimed to represent. In order for the “tree of life” to demonstrate the lines of evolution of species it must be established somehow that species can actually evolve like that.

    John Harshman: A question for you: do you think that there was one origin of language, so that all the world’s languages are related even though there’s no preserved evidence of it? Or do you think there are many independent origins of human language?

    Many independent origins. I believe this about species: Species emerged as a whole and whatever is irreducible to the species was there in them all along. Language is irreducible to humans and different groups of humans most likely had different languages from the very beginning. Just speculation, to be sure, not being a specialist (that is, not a specialist in how species emerge and not overly concerned with OOL – origin of language), but no worse than Darwin’s speculation whom you have openly discarded.

    John Harshman: Is there not disagreement on just where the trail stops?

    Given the same standard, no disagreement at all. Disagreements involve different standards. Nobody disagrees about the data. There are disagreements about the interpretation of the data and the interpretation depends on the methodological standard.

    It’s how science works, you know. You keep telling me that you know how it works.

    John Harshman: You create a strawman caricature of phylogenetics and then scoff at it. It’s not “some similarity or other”. It’s a huge set of similarities objectively assessed. Again, I wonder if you have ever read any of the phylogenetics literature after 1859. If you’re interested in the evidence for universal common descent specifically, you need to read Theobald 2010.

    Theobald D.L. A formal test of the theory of universal common ancestry. Nature 2010; 465:219-223.

    http://theobald.brandeis.edu/pdfs/Theobald_2010_Nature_all.pdf

    Thanks. I will get acquainted with genetics once I have the time. Meanwhile, I find it very telling that you as one familiar with genetics do not answer any of the basic questions. How do genes show what you claim they show, such as common descent? How is it common descent rather than the same building blocks used over and over?

    Will your answer be that the answers are found in the article? So the article is apparently unsummarizable. How convenient.

    And no, I did not create a strawman. I actually quoted Darwin who said “we have to make out community of descent by resemblances of any kind.” That’s his methodological standard, hopefully as dubious in biology as it would be in linguistics, so criticism is justified. Yet you fail to criticize him coherently and systematically, with examples. You only say, “It’s not “some similarity or other”. It’s a huge set of similarities objectively assessed.” No examples, just claims of objectivity. The problem is that Darwin claimed the same: Similarities a-plenty and I’m being objective here!

  18. Allan Miller: Now, I take a set of genes from a taxon. I build a tree on the pattern of identities in that set. ‘Cos, y’know, you can make a tree out of anything.

    Now do this with a completely different set of genes from the same taxon. Fuck me, it’s the same tree. Let’s take another. Nope, same. Even looking at ‘scratches’ – the kind of difference that makes no difference, and yes there really are such things – they arrange into the same tree.

    How so? It’s as if the vessel type, and colour, and maker’s marks, and blemishes, all group in the same way. Common descent says why. ‘I can arrange anything into a tree’ doesn’t.

    Compare:

    Rumraket: As I understand it trees can go anywhere from being highly incongruent in the sense that they don’t at all lend any support to a genealogical relationship, through a sort of grey zone where the results are just too uncertain to point one way or the other, to a match that would be perfectly congruent.

    These two views cannot be right at the same time concerning the same matter. I’m sure you guys can sort this out among yourselves. Best in the open with examples and evidence.

  19. Erik,

    These two views cannot be right at the same time concerning the same matter. I’m sure you guys can sort this out among yourselves.

    You are right, the trees in my illustration were an ideal. Due to many causes, the congruence of trees is rarely as perfect as my idealised illustration.

    Now, armed with that clarification, you can address the substance of the point, viz: trees built on separate characters should not even approximate the same tree outside of common descent.

  20. Allan Miller: Now, armed with that clarification, you can address the substance of the point, viz: trees built on separate characters should not even approximate the same tree outside of common descent.

    Sure, as soon as we have concrete examples. Let’s have a few examples where the trees drawn on separate data line up ideally and a few where they don’t.

    Without examples, it still looks very much like in linguistics: On broad strokes, we can draw a tree on anything and combine it into a worldwide tree. Given a more rigorous standard, the picture changes radically, the trees are several and cannot be combined across language families.

    After all these pages of comments, we should have already had some examples. Whatever, I will look them up myself when I have the time. Hopefully Theobald 2010 has some.

  21. Erik,

    Sure, as soon as we have concrete examples. Let’s have a few examples where the trees drawn on separate data line up ideally and a few where they don’t.

    We’ve had many examples already, such as John’s ratite paper and the earlier work it references. In fact, see the entire field of phylogenetic analysis. Not every study uses the same character set, and yet every study gives broadly congruent trees, give or take uncertainty at some nodes. Adding data resolves uncertainties, a most curious fact if common descent is not a thing among incompatible species.

    But let’s look at this:

    Whales and SINEs

    Whales and mitochondria

    There is no reason other than common descent to suppose that (fairly) congruent trees would be produced by mitochondrial genomes and by SINE inserts in nuclear genes***. I’m sure there are other studies using other genes, and I’m betting they are broadly congruent also. In fact that would be a test. I made that statement without looking. How can I be so sure? Lucky guess? Care to check?

    *** I know you thought you’d destroyed SINEs by finding an anomaly, but that anomaly only stood out at all because it was incongruent.

  22. Erik: Allan thinks I am feigning ignorance about genetics. In reality, I can’t make sense of absolutely anything you are talking about.

    We have identified the problem.

    Why do you think you fail to make sense of these things Erik?

  23. colewd: The experiment shows elevated vitamin C level in fetuses. The following paper shows that RNA editing can create a premature stop condon. .

    There is real experimental evidence here John, which is woefully missing in your citations such as the Theobald 2010 paper you cited that claims a probability calculation with no experimental evidence to support it.

    I can’t facepalm enough.
    I literally cannot.
    facepalm.
    enough.

  24. Allan Miller: We’ve had many examples already, such as John’s ratite paper and the earlier work it references. In fact, see the entire field of phylogenetic analysis. Not every study uses the same character set, and yet every study gives broadly congruent trees, give or take uncertainty at some nodes. Adding data resolves uncertainties, a most curious fact if common descent is not a thing among incompatible species.

    In other words, go read the papers. Nothing can be summarized (which kind of hints that you have not read anything yourself, or if you did, you understood nothing) and no answers to the basic questions ever. Got it already.

    Allan Miller: *** I know you thought you’d destroyed SINEs by finding an anomaly, but that anomaly only stood out at all because it was incongruent.

    And incongruences should matter as much as congruences. More importantly, their nature and causes matter. If there are more incongruences than there are congruences and the incongruences are insurmountable and inexplicable, it does not bode well for common descent.

  25. Erik: In other words, go read the papers.

    Yes. Why the FUCK should people bother spelling it all out for you here in posts, when it’s already assembled and compiled in accesible format in scientific publications?

    Nothing can be summarized

    Or maybe it can, but there’s a limit to how much you can expect others to do YOUR work for you.

    (which kind of hints that you have not read anything yourself, or if you did, you understood nothing)

    Or maybe that’s a false dichotomy for reasons already stated.

    and no answers to the basic questions ever.

    Every “basic” question of yours has been answered. But the problem is, in your own words, that “In reality, I can’t make sense of absolutely anything you are talking about.”

    Perhaps we should first work out why you are so curiously and oddly lacking in comprehension on this subject.

    And incongruences should matter as much as congruences. More importantly, their nature and causes matter. If there are more incongruences than there are congruences and the incongruences are insurmountable and inexplicable, it does not bode well for common descent.

    Well not exactly. There can in principle be more incongruent branches, than congruent branches, in a comparison between two phylogenetic trees, and yet they can still match with a high degree of statistical significant (read Theobald below). Nevertheless, in so far as there are incongruences, they are NOT insurmountable and inexplicable.

    Theobald wrote at length on this particular question:

    In science, independent measurements of theoretical values are never exact. When inferring any value (such as a physical constant like the charge of the electron, the mass of the proton, or the speed of light) some error always exists in the measurement, and all independent measurements are incongruent to some extent. Of course, the true value of something is never known for certain in science—all we have are measurements that we hope approximate the true value. Scientifically, then, the important relevant questions are “When comparing two measurements, how much of a discrepancy does it take to be a problem?” and “How close must the measurements be in order to give a strong confirmation?” Scientists answer these questions quantitatively with probability and statistics (Box 1978; Fisher 1990; Wadsworth 1997). To be scientifically rigorous we require statistical significance. Some measurements of a given value match with statistical significance (good), and some do not (bad), even though no measurements match exactly (reality).

    So, how well do phylogenetic trees from morphological studies match the trees made from independent molecular studies? There are over 10^38 different possible ways to arrange the 30 major taxa represented in Figure 1 into a phylogenetic tree (see Table 1.3.1; Felsenstein 1982; Li 1997, p. 102). In spite of these odds, the relationships given in Figure 1, as determined from morphological characters, are completely congruent with the relationships determined independently from cytochrome c molecular studies (for consensus phylogenies from pre-molecular studies see Carter 1954, Figure 1, p. 13; Dodson 1960, Figures 43, p. 125, and Figure 50, p. 150; Osborn 1918, Figure 42, p. 161; Haeckel 1898, p. 55; Gregory 1951, Fig. opposite title page; for phylogenies from the early cytochrome c studies see McLaughlin and Dayhoff 1973; Dickerson and Timkovich 1975, pp. 438-439). Speaking quantitatively, independent morphological and molecular measurements such as these have determined the standard phylogenetic tree, as shown in Figure 1, to better than 38 decimal places. This phenomenal corroboration of universal common descent is referred to as the “twin nested hierarchy”. This term is something of a misnomer, however, since there are in reality multiple nested hierarchies, independently determined from many sources of data.

    When two independently determined trees mismatch by some branches, they are called “incongruent”. In general, phylogenetic trees may be very incongruent and still match with an extremely high degree of statistical significance (Hendy et al. 1984; Penny et al. 1982; Penny and Hendy 1986; Steel and Penny 1993). Even for a phylogeny with a small number of organisms, the total number of possible trees is extremely large. For example, there are about a thousand different possible phylogenies for only six organisms; for nine organisms, there are millions of possible phylogenies; for 12 organisms, there are nearly 14 trillion different possible phylogenies (Table 1.3.1; Felsenstein 1982; Li 1997, p. 102). Thus, the probability of finding two similar trees by chance via two independent methods is extremely small in most cases. In fact, two different trees of 16 organisms that mismatch by as many as 10 branches still match with high statistical significance (Hendy et al. 1984, Table 4; Steel and Penny 1993). For more information on the statistical significance of trees that do not match exactly, see “Statistics of Incongruent Phylogenetic Trees“.

    The stunning degree of match between even the most incongruent phylogenetic trees found in the biological literature is widely unappreciated, mainly because most people (including many biologists) are unaware of the mathematics involved (Bryant et al. 2002; Penny et al. 1982; Penny and Hendy 1986). Penny and Hendy have performed a series of detailed statistical analyses of the significance of incongruent phylogenetic trees, and here is their conclusion:

    “Biologists seem to seek the ‘The One Tree’ and appear not to be satisfied by a range of options. However, there is no logical difficulty in having a range of trees. There are 34,459,425 possible [unrooted] trees for 11 taxa (Penny et al. 1982), and to reduce this to the order of 10-50 trees is analogous to an accuracy of measurement of approximately one part in 106.” (Penny and Hendy 1986, p. 414)

    I was so nice I even made sure the link to the statistics of incongruent phylogenies works.

    What’s next, you want us to feed you all the references in tiny digestible bites for you? In crayon cartoons?

  26. Erik,

    In other words, go read the papers. […]

    Well, yes. You would believe my summary? Would you hell as like.

    Allan Miller: *** I know you thought you’d destroyed SINEs by finding an anomaly, but that anomaly only stood out at all because it was incongruent.

    Erik: And incongruences should matter as much as congruences. More importantly, their nature and causes matter. If there are more incongruences than there are congruences and the incongruences are insurmountable and inexplicable, it does not bode well for common descent.

    If there are many more incongruences than congruences, you haven’t got a worthwhile mapping of trees. That’s not the case in the mouse SINE example.

    You’re right, where hypothetical data don’t support common descent, they don’t support common descent. Now, back too the stuff that does …

  27. Erik,

    (which kind of hints that you have not read anything yourself, or if you did, you understood nothing)

    That’s what you glean from my posting two papers? A master of the sneer, thou art, despite self-confessed ignorance in genetics.

  28. Rumraket,

    What’s next, you want us to feed you all the references in tiny digestible bites for you? In crayon cartoons?

    I think that’s it. Too many words.

  29. The dumb thing is, incongruences do matter as much as congruences. Where you have an incongruence, you have a piece of data demanding of an explanation. Further investigation can often uncover it. This is where data on gene transfers comes from. It sticks out from generally congruent trees like Pluto on a blink comparator. Without that background congruence, there’s nothing incongruent about it.

  30. Erik: How do genes show what you claim they show, such as common descent? How is it common descent rather than the same building blocks used over and over?

    I’m pretty sure this has been answered several times, but you have your hands over your eyes and ears. But I’ll try again. The patterns of differences among species are arranged in a nested hierarchy, and it’s the same nested hierarchy regardless of what data you choose. (Now of course that’s an oversimplification, but it applies very strongly within groups that aren’t too distantly related. Primates, say. Some phylogenetic problems are hard, while others are easy. Primates are the latter.)

    Why not the same building blocks used over and over? Because that doesn’t predict a nested hierarchy. It isn’t clear to me what you mean by “building blocks”. I suspect, being ignorant of genetics, you don’t know yourself. If you mean particular genes, then isn’t it odd that they aren’t the same building blocks, just rather similar ones, with the differences arranged in a nested hierarchy?

    And no, I did not create a strawman. I actually quoted Darwin who said “we have to make out community of descent by resemblances of any kind.” That’s his methodological standard, hopefully as dubious in biology as it would be in linguistics, so criticism is justified. Yet you fail to criticize him coherently and systematically, with examples. You only say, “It’s not “some similarity or other”. It’s a huge set of similarities objectively assessed.” No examples, just claims of objectivity. The problem is that Darwin claimed the same: Similarities a-plenty and I’m being objective here!

    I don’t think you are. You’re quote-mining. Tell me just where Darwin said that and I’ll look at what he meant.

    If you want examples, you really will have to read some of the scientific literature. The examples are big data sets with hundreds or thousands of characters. I can show you, but I can’t make you read. Here’s a comparatively small and simple example for you:

    Hayasaka K., Gojobori T., Horai S. Molecular phylogeny and evolution of primate mitochondrial DNA. Molecular Biology and Evolution 1988; 5:626-644.

  31. Speaking of Hayasaka K., Gojobori T., Horai S. Molecular phylogeny and evolution of primate mitochondrial DNA. Molecular Biology and Evolution 1988; 5:626-644, here’s a link to the summary and explanation you requested, one I wrote quite a few years ago:

    http://www.talkorigins.org/origins/postmonth/apr05.html

    It takes a small piece of the tree and looks at the evidence in detail. Added bonus for you, it also deals with data conflict.

  32. Rumraket: Yes. Why the FUCK should people bother spelling it all out for you here in posts, when it’s already assembled and compiled in accessible format in scientific publications?

    That’s it exactly. Erik strikes me as being fundamentally lazy. (Not that he’s unusual in that regard; most of the people at TSZ don’t want to do any of the hard work involved in reading philosophy.)

    What’s the point of trying to engage in dialogue with someone who refuses to learn anything about the topic you’re trying to talk about?

    One would have about as much success talking with Donald Trump about international trade or climate change as one would talking with Erik (or any of “theists” here) about biology.

  33. Kantian Naturalist: That’s it exactly. Erik strikes me as being fundamentally lazy.

    You mean when I have read Darwin, then Darwin gets rejected here by John Harshman, and I ask how biologists can still hold to common descent, without getting a sensible answer, it’s me being lazy?

    Yep, that’s what you mean. That’s KN.

  34. John Harshman: I’m pretty sure this has been answered several times, but you have your hands over your eyes and ears. But I’ll try again. The patterns of differences among species are arranged in a nested hierarchy, and it’s the same nested hierarchy regardless of what data you choose. (Now of course that’s an oversimplification, but it applies very strongly within groups that aren’t too distantly related. Primates, say. Some phylogenetic problems are hard, while others are easy. Primates are the latter.)

    And I’m pretty sure I have explained several times why this is the wrong answer. Of course it applies to more similar things, such as primates, “very strongly” that they can be arranged in a nested hierarchy by their similarities, while the same doesn’t apply to less similar things. The problem is in the assumption that despite all the problems with the more disparate things they still have common descent somehow. And you consider it proven with no need to acknowledge any even remotely lingering issue here. On what basis?

    John Harshman: Why not the same building blocks used over and over? Because that doesn’t predict a nested hierarchy.

    “Predict” in what sense? We have been over this point too. “Predict” is the wrong word here.

    Historical linguistics also “predicts” that languages are related so you should be able to draw a tree. The problem is that by lowering your bar you can demonstrate similarities between any two languages of your choice and call it a day. This is not even an issue. The issue is to keep things to a meaningful methodological standard.

    John Harshman: It isn’t clear to me what you mean by “building blocks”. I suspect, being ignorant of genetics, you don’t know yourself.

    I mean the same things you are looking at when you say they are similar so that they constitute homologies. Obviously. I suspect, willing to forever ignore the actual issues, you resort to talking past the issues.

    John Harshman: If you mean particular genes, then isn’t it odd that they aren’t the same building blocks, just rather similar ones, with the differences arranged in a nested hierarchy?

    You mean when you exemplified this by things like A and G, then an A in one animal or gene is not the same as A in another? Okay, you succeeded in obfuscating the issue irreparably. We shouldn’t stoop to debating the meanings of “same” and “similar”. Unless this is indeed the only way to prove common descent.

    John Harshman: Tell me just where Darwin said that and I’ll look at what he meant.

    I quoted it at length earlier. You didn’t look it up. Why would you now? Anyway, here http://www.gradesaver.com/the-origin-of-species/e-text/chapter-13-mutual-affinities-of-organic-beings-morphology-embryology-rudimentary-organs

    John Harshman: http://www.talkorigins.org/origins/postmonth/apr05.html

    Thanks. Quote from there, “Let’s try a statistical test just to be sure.” This is again the wrong answer and we have again been over this already. Statistics tells you nothing about what came from where. Does the same/similar statistical test on the genes of two random humans tell you that they are the same family versus not? I have seen you earlier say no. So what’s so drastically different here?

  35. Erik: How do genes show what you claim they show, such as common descent? How is it common descent rather than the same building blocks used over and over?

    How do you miss the point over and over? The “same building blocks” are used over and over in a derivative, hereditary way. You take the flying vertebrates, they all have “the same building blocks” of the tetrapod skeleton, they’re eukaryotes, amniotes, etc. This makes sense if they had a common ancestor, but maybe they were built using the same building blocks?

    But none of their flying adaptations have “the same building blocks.” Why not? Because the common ancestor of all of them was not a flying animal. Pterosaurs, birds, and bats all have modified tetrapod forelimbs into wings, none took “wing design” from any earlier (or later, as might be possible with a designer) organisms, each one simply modified limbs that had moved the animals across the ground at one time into wings.

    Evolution has no choice, it knows nothing about pterosaur wings when birds are evolving, or when bats are evolving. Intelligence does know other things, it is not limited like unthinking evolutionary processes are.

    Birds also have the good fortune of having excellent lungs that help to fuel their flight. But bats don’t get to use those lungs, they’re stuck with mammalian lungs. Birds also have better compensation for blood vessels being in front of the retina, the pecten. Do bats get that? No, of course not, they’re stuck with mammal eyes. But bats get the excellent mammalian ear bones, and birds do not, again because bats evolved from mammals and birds did not.

    The trees are just a way of illustrating the rather intense restraints that hereditary derivation that provides the mechanisms of evolution operate under. Evolution can only use “the same building blocks” over and over if it has access to these, and although there are cases of horizontal gene transfer, in the vertebrate lines you are primarily stuck with transmission down the hereditary lines. That’s why bats don’t get bird feathers and bird lungs, while birds don’t get the ear bones of the bat, or lactation.

    These are very basic facts that evolution explains and that nothing else does at all.

    Glen Davidson

  36. Erik: And no, I did not create a strawman. I actually quoted Darwin who said “we have to make out community of descent by resemblances of any kind.” That’s his methodological standard, hopefully as dubious in biology as it would be in linguistics, so criticism is justified.

    Quote-mining. The rest of the paragraph, starting directly after that quoted sentence, talks about what kind of resemblances — not just “any kind” — are useful. I believe I am required by the rules not to accuse you of dishonesty, but I think obliviousness is still available.

  37. Erik: You mean when I have read Darwin, then Darwin gets rejected here by John Harshman, and I ask how biologists can still hold to common descent, without getting a sensible answer, it’s me being lazy?

    Yep, that’s what you mean. That’s KN.

    More precisely: Darwin used an analogy to convey the concept of common descent, you found fault with that analogy (while ignoring all the evidence), and nothing that Harshman said made sense to you — ergo, all the fault lies on the part of evolutionary biologists and none on your own side.

    Right.

  38. Back to Darwin … sigh.

    Molecular sequences. There is no reason outside of common descent to expect a tree constructed using one set of sequences to be congruent – even ‘fairly congruent’ – with a tree based upon another. If one is unsure after two, one could pick a third. And so on.

  39. Erik: And I’m pretty sure I have explained several times why this is the wrong answer. Of course it applies to more similar things, such as primates, “very strongly” that they can be arranged in a nested hierarchy by their similarities, while the same doesn’t apply to less similar things. The problem is in the assumption that despite all the problems with the more disparate things they still have common descent somehow. And you consider it proven with no need to acknowledge any even remotely lingering issue here. On what basis?

    Whenever confronted on your opposition to common descent among restricted taxonomic groups, you try to change the subject to universal common descent. You really can’t get away with that when anyone is paying attention.

    “Predict” in what sense? We have been over this point too. “Predict” is the wrong word here.

    Why? By “predict” I mean that a nested hierarchy is exactly what we would expect to see given common descent.

    Historical linguistics also “predicts” that languages are related so you should be able to draw a tree. The problem is that by lowering your bar you can demonstrate similarities between any two languages of your choice and call it a day. This is not even an issue. The issue is to keep things to a meaningful methodological standard.

    Nobody is lowering any bars here. And “able to draw a tree” isn’t the criterion.

    I mean the same things you are looking at when you say they are similar so that they constitute homologies. Obviously.

    I don’t think you know what those things are. That’s one of your problems. You’re trying to dismiss a subject when you don’t even know what we’re talking about.

    You mean when you exemplified this by things like A and G, then an A in one animal or gene is not the same as A in another? Okay, you succeeded in obfuscating the issue irreparably. We shouldn’t stoop to debating the meanings of “same” and “similar”. Unless this is indeed the only way to prove common descent.

    I’m pretty sure that even at this point you have no idea what A and G are. We are not debating the meanings of “same” and “similar”. It’s just that you don’t know a gene from a base pair, have no interest in learning, and so can’t intelligently discuss the subject.

    Thanks. Quote from there, “Let’s try a statistical test just to be sure.” This is again the wrong answer and we have again been over this already. Statistics tells you nothing about what came from where. Does the same/similar statistical test on the genes of two random humans tell you that they are the same family versus not? I have seen you earlier say no. So what’s so drastically different here?

    You don’t know what you saw me say, because, again, you understand nothing of what you read. The discussion of random humans was in regard to bootstrapping, which bears no resemblance to a chi square test. I can certainly see how you could adapt a chi square to a paternity test, though again it shouldn’t be necessary given the number of loci generally involved and the unmistakable expected pattern.

    Now, what do statistics tell you? In the case I explained, they tell you that the distribution of states among species can’t be due to chance, and that’s all the chi square test was intended to do. Thus we must seek an explanation. Common descent explains the data. What else could explain the data? You have never even tried to suggest an alternative.

  40. John Harshman: By “predict” I mean that a nested hierarchy is exactly what we would expect to see given common descent.

    I wonder if it might be better to explain this in terms of what Charles Peirce called “abduction”.

    Peirce puts it as follows: “fact X is puzzling. But if Y were the case, then X would follow as a matter of course.”

    It’s not that common descent predicts nested hierarchies, but that common descent explains nested hierarchies. It’s the best abductive inference currently available to us. The underlying logic is “nested hierarchies are puzzling, but if there is common descent, then nested hierarchies would follow as a matter of course.”

  41. Kantian Naturalist: It’s not that common descent predicts nested hierarchies, but that common descent explains nested hierarchies.

    I’m not sure I see the distinction. “Would follow as a matter of course” sounds like a prediction to me. Or perhaps you would like “an expectation”. “A unique expectation”, even better.

Leave a Reply